Endoscopic image processing method and apparatus, electronic device, and storage medium

Abstract

This application provides an endoscopic image processing method and apparatus, an electronic device, and a storage medium. The method includes: obtaining an endoscopic video stream, the endoscopic video stream including an original endoscopic image; detecting the original endoscopic image in a corresponding video frame through a first thread, and transmitting a detection result of the original endoscopic image to an integration module; forming a control instruction through the integration module according to the detection result; and adjusting, in response to the control instruction and through a second thread, an output result in the second thread, the output result corresponding to a use environment of the endoscopic video stream, the first thread and the second thread being parallel threads.

What is claimed is: 
     
       1. An endoscopic image processing method, performed by an electronic device, the endoscopic image processing method comprising:
 obtaining an endoscopic video stream, the endoscopic video stream including an original endoscopic image; 
 detecting the original endoscopic image in a corresponding video frame through a first thread, and transmitting a detection result of the original endoscopic image to an integration module, comprising:
 detecting a foreign body through a detector in the first thread to obtain a foreign body box distributed in a target endoscopic image frame, the target endoscopic image frame being obtained based on the original endoscopic image, and the foreign body box indicating a region where the foreign body exists in the target endoscopic image frame; 
 filtering the target endoscopic image frame based on the foreign body box, to position a lesion region in the target endoscopic image frame and determine a lesion category; and 
 transmitting the lesion region in the target endoscopic image frame to the integration module; 
 
 forming a control instruction through the integration module according to the detection result; and 
 adjusting, in response to the control instruction and through a second thread, an output result in the second thread, the output result corresponding to a use environment of the endoscopic video stream, the first thread and the second thread being parallel threads. 
 
     
     
       2. The method according to  claim 1 , wherein the obtaining an endoscopic video stream comprises:
 extracting an endoscopic video frame in pathological information of a target object; 
 improving resolution of the endoscopic video frame; 
 converting a format of the endoscopic video frame from a current encoding format to a grayscale value encoding format, to obtain a plurality of endoscopic video frames; and 
 performing encoding compression on the endoscopic video frame to form an endoscopic video stream. 
 
     
     
       3. The method according to  claim 2 , wherein the converting a format of the endoscopic video frame from a current encoding format to a grayscale value encoding format comprises:
 performing encoding compression on the endoscopic video frame using an encoding method corresponding to a target format; 
 writing a serial number corresponding to the video frame and generating a timestamp corresponding to an image packet, to generate a packet corresponding to the format of the endoscopic video frame; and 
 splicing the image packet to form a plurality of endoscopic video frames. 
 
     
     
       4. The method according to  claim 1 , wherein the method further comprises:
 adjusting the original endoscopic image to a target size to generate standard endoscopic image frames; and 
 filtering the standard endoscopic image frames and deleting a standard endoscopic image frame having interference, to form the target endoscopic image frame comprising different target endoscopic image frames. 
 
     
     
       5. The method according to  claim 1 , wherein the method further comprises:
 when a detection time of the detector in the first thread exceeds a detection time threshold, stopping detecting a current target endoscopic image frame, and obtaining a subsequent target endoscopic image frame for detection. 
 
     
     
       6. The method according to  claim 1 , wherein the adjusting, in response to the control instruction and through a second thread, an output result in the second thread comprises:
 when a tracker in the second thread is in a continuation state and a tracking box corresponding to the tracker matches a foreign body box corresponding to a detector, outputting, by the tracker in the second thread, a current target endoscopic image frame; or 
 when a tracker in the second thread is in a re-initiation state and a tracking box corresponding to the tracker does not match a foreign body box corresponding to a detector, outputting, by the tracker in the second thread, a new target endoscopic image frame. 
 
     
     
       7. The method according to  claim 6 , wherein the method further comprises:
 determining a corresponding box regression parameter when a foreign body box exists in the target endoscopic image frame; and 
 activating the tracker in the second thread by using the box regression parameter. 
 
     
     
       8. The method according to  claim 7 , wherein the method further comprises:
 when the tracker in the second thread is activated, distributing a target identifier to a foreign body in the current target endoscopic image frame; and 
 when foreign bodies in target endoscopic image frames when the tracker in the second thread is in an activated state, a continuation state, and a re-initiation state are consistent, initializing the target identifier to implement continuous tracking on the foreign body in the current target endoscopic image frame. 
 
     
     
       9. The method according to  claim 1 , wherein the method further comprises:
 transmitting an adjustment instruction in response to the output result in the second thread, the adjustment instruction adjusting a detection state of a medical device in contact with a target object, to obtain new pathological information. 
 
     
     
       10. An electronic device, comprising:
 a memory, configured to store executable instructions; and 
 a processor, configured to, when executing the executable instructions stored in the memory, perform: 
 obtaining an endoscopic video stream, the endoscopic video stream including an original endoscopic image; 
 detecting the original endoscopic image in a corresponding video frame through a first thread, and transmitting a detection result of the original endoscopic image to an integration module comprising:
 detecting a foreign body through a detector in the first thread to obtain a foreign body box distributed in a target endoscopic image frame, the target endoscopic image frame being obtained based on the original endoscopic image, and the foreign body box indicating a region where the foreign body exists in the target endoscopic image frame; 
 filtering the target endoscopic image frame based on the foreign body box, to position a lesion region in the target endoscopic image frame and determine a lesion category; and 
 transmitting the lesion region in the target endoscopic image frame to the integration module; 
 
 forming a control instruction through the integration module according to the detection result; and 
 adjusting, in response to the control instruction and through a second thread, an output result in the second thread, the output result corresponding to a use environment of the endoscopic video stream, the first thread and the second thread being parallel threads. 
 
     
     
       11. A non-transitory computer-readable storage medium, storing executable instructions, the executable instructions, when executed by a processor, implementing:
 obtaining an endoscopic video stream, the endoscopic video stream including an original endoscopic image; 
 detecting the original endoscopic image in a corresponding video frame through a first thread, and transmitting a detection result of the original endoscopic image to an integration module comprising:
 detecting a foreign body through a detector in the first thread to obtain a foreign body box distributed in a target endoscopic image frame, the target endoscopic image frame being obtained based on the original endoscopic image, and the foreign body box indicating a region where the foreign body exists in the target endoscopic image frame; 
 filtering the target endoscopic image frame based on the foreign body box, to position a lesion region in the target endoscopic image frame and determine a lesion category; and 
 transmitting the lesion region in the target endoscopic image frame to the integration module; 
 
 forming a control instruction through the integration module according to the detection result; and 
 adjusting, in response to the control instruction and through a second thread, an output result in the second thread, the output result corresponding to a use environment of the endoscopic video stream, the first thread and the second thread being parallel threads. 
 
     
     
       12. The computer-readable storage medium according to  claim 11 , wherein the obtaining an endoscopic video stream comprises:
 extracting an endoscopic video frame in pathological information of a target object; 
 improving resolution of the endoscopic video frame; 
 converting a format of the endoscopic video frame from a current encoding format to a grayscale value encoding format, to obtain a plurality of endoscopic video frames; and 
 performing encoding compression on the endoscopic video frame to form an endoscopic video stream. 
 
     
     
       13. The computer-readable storage medium according to  claim 12 , wherein the converting a format of the endoscopic video frame from a current encoding format to a grayscale value encoding format comprises:
 performing encoding compression on the endoscopic video frame using an encoding method corresponding to a target format; 
 writing a serial number corresponding to the video frame and generating a timestamp corresponding to an image packet, to generate a packet corresponding to the format of the endoscopic video frame; and 
 splicing the image packet to form a plurality of endoscopic video frames. 
 
     
     
       14. The computer-readable storage medium according to  claim 11 , wherein the executable instructions further causes the processor to implement:
 adjusting the original endoscopic image to a target size to generate standard endoscopic image frames; and 
 filtering the standard endoscopic image frames and deleting a standard endoscopic image frame having interference, to form the target endoscopic image frame comprising different target endoscopic image frames. 
 
     
     
       15. The computer-readable storage medium according to  claim 11 , wherein the executable instructions further causes the processor to implement:
 when a detection time of the detector in the first thread exceeds a detection time threshold, stopping detecting a current target endoscopic image frame, and obtaining a subsequent target endoscopic image frame for detection. 
 
     
     
       16. The computer-readable storage medium according to  claim 11 , wherein the adjusting, in response to the control instruction and through a second thread, an output result in the second thread comprises:
 when a tracker in the second thread is in a continuation state and a tracking box corresponding to the tracker matches a foreign body box corresponding to a detector, outputting, by the tracker in the second thread, a current target endoscopic image frame; or 
 when a tracker in the second thread is in a re-initiation state and a tracking box corresponding to the tracker does not match a foreign body box corresponding to a detector, outputting, by the tracker in the second thread, a new target endoscopic image frame. 
 
     
     
       17. The computer-readable storage medium according to  claim 16 , wherein the executable instructions further causes the processor to implement:
 determining a corresponding box regression parameter when a foreign body box exists in the target endoscopic image frame; and 
 activating the tracker in the second thread by using the box regression parameter. 
 
     
     
       18. The computer-readable storage medium according to  claim 17 , wherein the executable instructions further causes the processor to implement:
 when the tracker in the second thread is activated, distributing a target identifier to a foreign body in the current target endoscopic image frame; and 
 when foreign bodies in target endoscopic image frames when the tracker in the second thread is in an activated state, a continuation state, and a re-initiation state are consistent, initializing the target identifier to implement continuous tracking on the foreign body in the current target endoscopic image frame.